From nobody Fri Feb 13 12:29:17 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id DE0B9CE7A8C for ; Mon, 25 Sep 2023 07:22:32 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S231536AbjIYHWg (ORCPT ); Mon, 25 Sep 2023 03:22:36 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:43526 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230245AbjIYHWb (ORCPT ); Mon, 25 Sep 2023 03:22:31 -0400 Received: from out-192.mta1.migadu.com (out-192.mta1.migadu.com [95.215.58.192]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 6214810D for ; Mon, 25 Sep 2023 00:22:21 -0700 (PDT) X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1695626539; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=Ap9IFexPW6VSs4Zh393WapxdcxvLLzkRsnjfydkgn14=; b=oWZzBjTqSTsx/0+SzwtdysDcYfGjdzBEi1R4gwWgDhGCgM41rNFA3mAzsEL7wvoEofACNn oMuG7DdU8H6ETVNhQPl3bURePR9gV1bmd7YubvPMYOV4qF8g8FyNQV++PMvyV2jPTXrXw3 idhBj/s4YZ1VfE26hx6ol8MWNsrvfjU= From: Yajun Deng To: akpm@linux-foundation.org, rppt@kernel.org Cc: mike.kravetz@oracle.com, muchun.song@linux.dev, willy@infradead.org, david@redhat.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org, Yajun Deng Subject: [PATCH v2 1/2] mm: pass page count and reserved to __init_single_page Date: Mon, 25 Sep 2023 15:21:49 +0800 Message-Id: <20230925072150.386880-2-yajun.deng@linux.dev> In-Reply-To: <20230925072150.386880-1-yajun.deng@linux.dev> References: <20230925072150.386880-1-yajun.deng@linux.dev> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" When we init a single page, we need to mark this page reserved if it does. And some pages may not need to set page count, such as compound pages. Introduce INIT_PAGE_COUNT and INIT_PAGE_RESERVED, let the caller decide if it needs to set page count and mark page reserved or not. Signed-off-by: Yajun Deng --- mm/hugetlb.c | 2 +- mm/internal.h | 8 +++++++- mm/mm_init.c | 31 +++++++++++++++++-------------- 3 files changed, 25 insertions(+), 16 deletions(-) diff --git a/mm/hugetlb.c b/mm/hugetlb.c index e2123d1bb4a2..b63635abe5c3 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -3196,7 +3196,7 @@ static void __init hugetlb_folio_init_tail_vmemmap(st= ruct folio *folio, for (pfn =3D head_pfn + start_page_number; pfn < end_pfn; pfn++) { struct page *page =3D pfn_to_page(pfn); =20 - __init_single_page(page, pfn, zone, nid); + __init_single_page(page, pfn, zone, nid, INIT_PAGE_COUNT); prep_compound_tail((struct page *)folio, pfn - head_pfn); ret =3D page_ref_freeze(page, 1); VM_BUG_ON(!ret); diff --git a/mm/internal.h b/mm/internal.h index 7a961d12b088..e9366cce461c 100644 --- a/mm/internal.h +++ b/mm/internal.h @@ -1209,8 +1209,14 @@ struct vma_prepare { struct vm_area_struct *remove2; }; =20 +enum { + INIT_PAGE_COUNT =3D (1 << 0), + INIT_PAGE_RESERVED =3D (1 << 1), +}; + void __meminit __init_single_page(struct page *page, unsigned long pfn, - unsigned long zone, int nid); + unsigned long zone, int nid, + unsigned int flags); =20 /* shrinker related functions */ unsigned long shrink_slab(gfp_t gfp_mask, int nid, struct mem_cgroup *memc= g, diff --git a/mm/mm_init.c b/mm/mm_init.c index 06a72c223bce..61df37133331 100644 --- a/mm/mm_init.c +++ b/mm/mm_init.c @@ -557,11 +557,11 @@ static void __init find_zone_movable_pfns_for_nodes(v= oid) } =20 void __meminit __init_single_page(struct page *page, unsigned long pfn, - unsigned long zone, int nid) + unsigned long zone, int nid, + unsigned int flags) { mm_zero_struct_page(page); set_page_links(page, zone, nid, pfn); - init_page_count(page); page_mapcount_reset(page); page_cpupid_reset_last(page); page_kasan_tag_reset(page); @@ -572,6 +572,10 @@ void __meminit __init_single_page(struct page *page, u= nsigned long pfn, if (!is_highmem_idx(zone)) set_page_address(page, __va(pfn << PAGE_SHIFT)); #endif + if (flags & INIT_PAGE_COUNT) + init_page_count(page); + if (flags & INIT_PAGE_RESERVED) + __SetPageReserved(page); } =20 #ifdef CONFIG_NUMA @@ -714,7 +718,7 @@ static void __meminit init_reserved_page(unsigned long = pfn, int nid) if (zone_spans_pfn(zone, pfn)) break; } - __init_single_page(pfn_to_page(pfn), pfn, zid, nid); + __init_single_page(pfn_to_page(pfn), pfn, zid, nid, INIT_PAGE_COUNT); } #else static inline void pgdat_set_deferred_range(pg_data_t *pgdat) {} @@ -821,8 +825,8 @@ static void __init init_unavailable_range(unsigned long= spfn, pfn =3D pageblock_end_pfn(pfn) - 1; continue; } - __init_single_page(pfn_to_page(pfn), pfn, zone, node); - __SetPageReserved(pfn_to_page(pfn)); + __init_single_page(pfn_to_page(pfn), pfn, zone, node, + INIT_PAGE_COUNT | INIT_PAGE_RESERVED); pgcnt++; } =20 @@ -884,7 +888,7 @@ void __meminit memmap_init_range(unsigned long size, in= t nid, unsigned long zone } =20 page =3D pfn_to_page(pfn); - __init_single_page(page, pfn, zone, nid); + __init_single_page(page, pfn, zone, nid, INIT_PAGE_COUNT); if (context =3D=3D MEMINIT_HOTPLUG) __SetPageReserved(page); =20 @@ -965,11 +969,9 @@ static void __init memmap_init(void) #ifdef CONFIG_ZONE_DEVICE static void __ref __init_zone_device_page(struct page *page, unsigned long= pfn, unsigned long zone_idx, int nid, - struct dev_pagemap *pgmap) + struct dev_pagemap *pgmap, + unsigned int flags) { - - __init_single_page(page, pfn, zone_idx, nid); - /* * Mark page reserved as it will need to wait for onlining * phase for it to be fully associated with a zone. @@ -977,7 +979,7 @@ static void __ref __init_zone_device_page(struct page *= page, unsigned long pfn, * We can use the non-atomic __set_bit operation for setting * the flag as we are still initializing the pages. */ - __SetPageReserved(page); + __init_single_page(page, pfn, zone_idx, nid, flags | INIT_PAGE_RESERVED); =20 /* * ZONE_DEVICE pages union ->lru with a ->pgmap back pointer @@ -1041,7 +1043,7 @@ static void __ref memmap_init_compound(struct page *h= ead, for (pfn =3D head_pfn + 1; pfn < end_pfn; pfn++) { struct page *page =3D pfn_to_page(pfn); =20 - __init_zone_device_page(page, pfn, zone_idx, nid, pgmap); + __init_zone_device_page(page, pfn, zone_idx, nid, pgmap, 0); prep_compound_tail(head, pfn - head_pfn); set_page_count(page, 0); =20 @@ -1084,7 +1086,8 @@ void __ref memmap_init_zone_device(struct zone *zone, for (pfn =3D start_pfn; pfn < end_pfn; pfn +=3D pfns_per_compound) { struct page *page =3D pfn_to_page(pfn); =20 - __init_zone_device_page(page, pfn, zone_idx, nid, pgmap); + __init_zone_device_page(page, pfn, zone_idx, nid, pgmap, + INIT_PAGE_COUNT); =20 if (pfns_per_compound =3D=3D 1) continue; @@ -2058,7 +2061,7 @@ static unsigned long __init deferred_init_pages(stru= ct zone *zone, } else { page++; } - __init_single_page(page, pfn, zid, nid); + __init_single_page(page, pfn, zid, nid, INIT_PAGE_COUNT); nr_pages++; } return (nr_pages); --=20 2.25.1 From nobody Fri Feb 13 12:29:17 2026 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3CDCECE7A81 for ; Mon, 25 Sep 2023 07:22:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230177AbjIYHWz (ORCPT ); Mon, 25 Sep 2023 03:22:55 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:48414 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S231896AbjIYHWl (ORCPT ); Mon, 25 Sep 2023 03:22:41 -0400 Received: from out-200.mta1.migadu.com (out-200.mta1.migadu.com [95.215.58.200]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 0C5A619F for ; Mon, 25 Sep 2023 00:22:27 -0700 (PDT) X-Report-Abuse: Please report any abuse attempt to abuse@migadu.com and include these headers. DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux.dev; s=key1; t=1695626546; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=u7hdNMC4UOcW2t7uwYGgTfN9WDCRbod/rNSZ3lnlhtw=; b=BXlLx+C48qOjacmeTMWRU5jZhJUdcFTtpDZuI2ajVt4e0J0SA08MBhucH4QPbVjAKyPmun h0Z0jblLvBQUT5O2gVLrp8WKrOnPSG/ynQtMXnmkDDf/KGlSUKhtGechJ/dINHbRNQC22I mgTeKnYclWq7yaIPfhAd3gFi7RMWn3E= From: Yajun Deng To: akpm@linux-foundation.org, rppt@kernel.org Cc: mike.kravetz@oracle.com, muchun.song@linux.dev, willy@infradead.org, david@redhat.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org, Yajun Deng Subject: [PATCH v2 2/2] mm: Init page count in reserve_bootmem_region when MEMINIT_EARLY Date: Mon, 25 Sep 2023 15:21:50 +0800 Message-Id: <20230925072150.386880-3-yajun.deng@linux.dev> In-Reply-To: <20230925072150.386880-1-yajun.deng@linux.dev> References: <20230925072150.386880-1-yajun.deng@linux.dev> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-Migadu-Flow: FLOW_OUT Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" memmap_init_range() would set page count of all pages, but the free pages count would be reset in __free_pages_core(). There are opposite operations. It's unnecessary and time-consuming when it's MEMINIT_EARLY context. Init page count in reserve_bootmem_region when in MEMINIT_EARLY context, and check the page count before reset it. At the same time, the INIT_LIST_HEAD in reserve_bootmem_region isn't need, as it already done in __init_single_page. The following data was tested on an x86 machine with 190GB of RAM. before: free_low_memory_core_early() 341ms after: free_low_memory_core_early() 285ms Signed-off-by: Yajun Deng --- mm/mm_init.c | 18 +++++++++++++----- mm/page_alloc.c | 20 ++++++++++++-------- 2 files changed, 25 insertions(+), 13 deletions(-) diff --git a/mm/mm_init.c b/mm/mm_init.c index 61df37133331..64c00ebaf4ef 100644 --- a/mm/mm_init.c +++ b/mm/mm_init.c @@ -718,7 +718,7 @@ static void __meminit init_reserved_page(unsigned long = pfn, int nid) if (zone_spans_pfn(zone, pfn)) break; } - __init_single_page(pfn_to_page(pfn), pfn, zid, nid, INIT_PAGE_COUNT); + __init_single_page(pfn_to_page(pfn), pfn, zid, nid, 0); } #else static inline void pgdat_set_deferred_range(pg_data_t *pgdat) {} @@ -756,8 +756,8 @@ void __meminit reserve_bootmem_region(phys_addr_t start, =20 init_reserved_page(start_pfn, nid); =20 - /* Avoid false-positive PageTail() */ - INIT_LIST_HEAD(&page->lru); + /* Set page count for the reserve region */ + init_page_count(page); =20 /* * no need for atomic set_bit because the struct @@ -888,9 +888,17 @@ void __meminit memmap_init_range(unsigned long size, i= nt nid, unsigned long zone } =20 page =3D pfn_to_page(pfn); - __init_single_page(page, pfn, zone, nid, INIT_PAGE_COUNT); - if (context =3D=3D MEMINIT_HOTPLUG) + + /* If the context is MEMINIT_EARLY, we will set page count and + * mark page reserved in reserve_bootmem_region, the free region + * wouldn't have page count and we will check the pages count + * in __free_pages_core. + */ + __init_single_page(page, pfn, zone, nid, 0); + if (context =3D=3D MEMINIT_HOTPLUG) { + init_page_count(page); __SetPageReserved(page); + } =20 /* * Usually, we want to mark the pageblock MIGRATE_MOVABLE, diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 06be8821d833..b868caabe8dc 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1285,18 +1285,22 @@ void __free_pages_core(struct page *page, unsigned = int order) unsigned int loop; =20 /* - * When initializing the memmap, __init_single_page() sets the refcount - * of all pages to 1 ("allocated"/"not free"). We have to set the - * refcount of all involved pages to 0. + * When initializing the memmap, memmap_init_range sets the refcount + * of all pages to 1 ("reserved" and "free") in hotplug context. We + * have to set the refcount of all involved pages to 0. Otherwise, + * we don't do it, as reserve_bootmem_region only set the refcount on + * reserve region ("reserved") in early context. */ - prefetchw(p); - for (loop =3D 0; loop < (nr_pages - 1); loop++, p++) { - prefetchw(p + 1); + if (page_count(page)) { + prefetchw(p); + for (loop =3D 0; loop < (nr_pages - 1); loop++, p++) { + prefetchw(p + 1); + __ClearPageReserved(p); + set_page_count(p, 0); + } __ClearPageReserved(p); set_page_count(p, 0); } - __ClearPageReserved(p); - set_page_count(p, 0); =20 atomic_long_add(nr_pages, &page_zone(page)->managed_pages); =20 --=20 2.25.1